• FDA Should Be Required To Provide 510(k) Decision Summaries
  • June 19, 2012 | Author: Jeffrey K. Shapiro
  • Law Firm: Hyman, Phelps & McNamara, P.C. - Washington Office
  • A provision in the House FDA Reform Act of 2012 (H.R. 5651) would require FDA to “regularly publish detailed decision summaries for each clearance of a device requiring clinical data” (Section 704). The requirement would be applied only on a going forward basis to new clearances. This provision is not in the companion Senate Food and Drug Administration Safety and Innovation Act (S. 3187). An earlier Senate proposal (S. 2292) did have this provision, and would have applied it to all clearances, not just those requiring clinical data. That bill, however, did not move forward to Senate passage.

    The House Report for H.R. 5651 says that Section 704 “would add predictability, consistency, and transparency to the medical device premarket review process.” House Report 112-495 at 26. Specifically, the Report states: “Only by creating transparency in decision making can we ensure the review process functions as intended.” Id. at 27. The House Report does not elaborate on why this reform would improve the 510(k) system, but it is worth considering this point in more detail.

    In each 510(k) review, FDA compares a proposed device to an accepted baseline, which is a “predicate” device generally already cleared to market via a previous 510(k) decision. The proposed device and the predicate must have the same intended use, but they need not have the same technological characteristics. The technological differences can be bridged if FDA finds that, based upon the data and information presented, the proposed device implicates the same type of safety and effectiveness questions as the predicate device, and is at least as safe and effective. This finding is known as “substantial equivalence.”

    The 510(k) system, then, consists of a series of case by case adjudications of individual device safety and effectiveness, each one based upon the substantial equivalence comparison just described. Once a new technology (or technological feature) has received clearance, it then becomes a platform for further incremental improvements. The system has proven to be adaptable to an almost infinite variety of device technologies, allowing FDA to conduct a meaningful premarket review of safety and effectiveness while still allowing continuous technological innovation.

    A persistent criticism of FDA’s implementation of the 510(k) system, however, is a lack of predictability, particularly as to the data that will be required to bridge the gap between proposed and predicate devices. The lack of predictability leads to delays and friction in the review process.

    A possible root cause of this problem actually may be a lack of transparency. Specifically, in order for the 510(k) system to operate predictably, the public must have access to the essential information in prior 510(k) decisions, such as the intended use and technological characteristics of the proposed and predicate devices, and the data required to show substantial equivalence. This information is necessary to allow a sponsor seeking clearance of a proposed device to hunt for predicate devices, learn FDA’s existing data requirements, and/or extrapolate likely additional data requirements that might apply.

    Unfortunately, the essential information is not easily available. The database of 510(k) summaries is the primary searchable database, and these summaries are wholly inadequate. They are prepared by the submitter, not FDA, and they are (i) not standardized very well and (ii) tend to be vague and lacking in detail (sometimes intentionally so, to avoid providing potential competitors with useful information about how to achieve clearance). FDA has forced improvements in the quality of the 510(k) summaries in the past couple of years, but not enough has been done. It is also impractical for the public to obtain a complete 510(k) file; a Freedom of Information Act request is required and the fulfillment time is typically a year or more.

    The bottom line is that the public is operating from a different and degraded database as compared to FDA reviewers, who have access to complete 510(k) files and also have institutional knowledge of their prior decisions. The current lack of access to important decision data creates a cloudy picture that leads to unpleasant surprises when FDA reviewers impose data requirements that were not discernible from the public 510(k) summaries.

    Section 704 would provide at least a partial remedy to this problem. As the House and Senate conferees reconcile their companion bills, they should certainly include section 704 of the House bill. However, section 704 is really too timid. It should be extended to require decision summaries for all new 510(k) clearances, as was proposed in S. 2292, not just the approximately 10%requiring clinical data. Indeed, the use of decision summaries is not new. It was implemented almost ten years ago by the Office of In Vitro Diagnostics (“OIVD”). As stated in OIVD’s FY 2005 Annual Report (P. 28):

    In an effort to provide stakeholders with the scientific / regulatory basis for FDA’s decisions . . . OIVD . . . implemented the use of a standardized Premarket Decision Summary Template across all the OIVD divisions.

    The Premarket Decision Summary Template summarizes the basis on which an in vitro diagnostic device was cleared under a 510(k) submission. OIVD implemented the use of this standardized premarket Decision Summary Template in August 1, 2003. The decision summaries have been continuously posted on the OIVD webpage and the public has full access to them through the 510(k) database.

    OIVD has continued posting decision summaries to this day. Congress should simply require FDA to extend this useful practice to the Office of Device Evaluation (“ODE”). If ODE lacks sufficient resources, then Congress should provide the necessary funding. However, it is not obvious that such funding is needed. The ODE reviewers undoubtedly document their review decisions now. It should not require significant additional resources to create a standard template, complete it for each 510(k) review, and post it to the 510(k) database. The experience of OIVD should be a guide as to what level of resources will be required.